兰州大学光转换材料与技术国家地方联合工程
实验室学术报告——Ranjan Srivastava教授

日期: 2018-08-07 阅读: 来源: 关键词:

  应兰州大学物理科学与技术学院和光转换材料与技术国家地方联合工程实验室邀请,美国康涅狄格大学Ranjan Srivastava教授将于8月8-11日间来我校访问并面向全校师生及功能与环境材料研究所做学术报告,敬请关注并欢迎大家参加!

  报告题目:Harnessing “Natural” Algorithms For Resolving Problems In Molecular and Biomolecular Engineering
  时   间:2018年8月10日(星期五)15:00-16:30
  地   点:格致楼3016会议室 

Ranjan Srivastava教授简介:

  Ranjan Srivastava, 2016年至今为美国康涅狄格大学化学和生物分子工程系教授、系主任;生物医学工程系教授;化学与生物分子系教授,2014年任耶鲁大学公共卫生学院生物统计学副教授,2002-2008年,为康涅狄格大学化学工程系助理教授。2000获得美国国家卫生研究院癌症生物学博士后学位,2004获罗杰斯杰出教学奖,2006年成为国家工程学院/医学院初级教师疫苗会议研究员,2009获得教学优秀奖,2013获康涅狄格大学工程学院杰出教师指导奖,2017年康涅狄格大学家庭支持奖的首名获奖者。

摘  要:

Nature has given rise to a variety of algorithms resulting in the development and evolution of populations of objects, entities, and even processes of extraordinary sophistication and complexity. If one can understand the basis on which these “natural” algorithms operate, the potential exists to harness this force. Such knowledge can provide us with a better understanding of the world around us. Additionally, one could use the insights gleaned to design and engineer items ranging from electronic devices to novel materials to recombinant organisms. The potential for these algorithms to be applied in this way has been recognized computationally in the form of artificial neural networks, particle swarm optimization, and evolutionary algorithms such as genetic algorithms to name just a few examples.  Our group has attempted to take advantage of these natural algorithms in a variety of ways. Over the course of this talk, I will share some of our efforts in two areas.  The first area is that of biomolecular engineering focusing on the use of genetic algorithms in genome-scale metabolic engineering.  Specifically, we show how a genetic algorithm may be used to curate metabolic models.  We will further go on to show how, by using such an approach, we were able to discover a new gene.  The second area will focus molecular engineering, specifically computational analysis and design of materials.  Here we will discuss how adding chemical descriptors in machine learning algorithms to characterize metal organic frameworks resulted in significant increase in accuracy of predicting material properties while reducing simulation time by several orders of magnitude.

发现错误?报错
文:
图:
编辑:潘纯熙
 

推荐关注

阅读下一篇